High idle at speed, not when stationary

I recently swapped my engine (it's a 996 Carrera), and got a strange problem. It idles high, at 1500 rpm whenever I clutch (with accompanying poor mileage). It will continue to stay at 1500 rpm, making me look like an idiot around town, until the car has been standing still for 10-20 seconds. It will then idle down to approx 900 rpm like it should. I swapped over the throttle-body from the old engine and it worked perfectly before. I checked the throttle valve and cleaned it out, but it did close properly already. Tried to clean the ICV as best I could, but without liquids. I did notice a clicking-sound from it that I'm not sure I've heard before? It may be related? Any ideas?

Quotes source post: 9768927

Even a 900 rpm idle speed is too high unless the engine is cold. Then idle speed will initially be, can be a bit above 1K rpms though it drops as the engine idles and gains heat. Next, swapped what MY engine into what MY Carrera? If you strayed too far away there can be issues, though someone I believe has resolved them. There's a Pelican Parts web site: www.pelicanparts.com that has some write ups covering engine 986 to 986 and 996 to 986 engine swaps -- not sure if there are any writes up covering 996 3.6l to say 996 3.4l car swaps though, and these write ups cover some of the issues involved. I haven't memorized this info so all I can suggest is find the site and read away. Off the top of my head, if the MY engine came with an ICV that would be my first suspect. A proper cleaning at least, followed by a check out to make sure the thing now clean works properly is called for. Sincerely, Macster.

Swapped an M96.01 engine for an M96.01 engine in a '98 c2. I wasn't looking to upgrade or anything, as most engines are a POS, and GT3/Turbo is too expensive. Besides, getting an hp-increase road-legal over here is a nightmare. As I wrote, the whole throttlebody from my working engine (before the cylinders became oval and cracked), was swapped over to another identical working engine. Reason being that the receptacle that holds the gas wire was snapped off on the "new" (used) engine and the easiest fix was just to move the whole thing over with TPS and ICV attached. It worked fine before so... The throttle butterfly itself wasn't sparkling clean but not that bad either. I wiped it down eitherway. Same with the ICV-valve, cleaned without liquids. But perhaps I could spray the ICV with carb-cleaner?

Quotes source post: 9770526

If I read you right you swapped like engine for like engine so the swap should be as bereft of any issues as any engine swap can be. Try the simplest things first. A proper cleaning of the ICV is a start. Be sure the engine install didn't damage/pinch any gas lines. While a fuel supply/pressure problem doesn't necessarily explain the idle speed issues fuel lines are always at risk of getting damaged/pinches in an engine swap. Be sure all vacuum hoses are ok and connected properly. Generally my training was at engine swap time to replace/renew all vacuum hoses with new hose material. Let's see... throttle control is by cable, yes? All you can do in this case is make sure the cable tension is adjusted right and the routing is factory. As with fuel hoses a throttle cable is at risk of being damaged when one's focus is on moving into position/securing a 400lb+ lump of engine. Be sure when the engine is first started, and probably twists some in/on its mounts, it doesn't tug on the throttle cable. In fact be sure the engine is secure and doesn't have a loose or even broken (the rubber mainly) mount. With an OBD2 code reader/data viewer check the intake and coolant temps at cold engine start along with the ignition advance to see if you can spot anything out of the ordinary, like temp readings that differ too much from each other or from ambient or timing that fluctuates. View the short term fuel trims and 02 sensor voltages looking for anything out of the ordinary from these readings that might point you in the right direction. Sincerely, Macster.

Yeah, I was trying to keep it simple, identical engines should be easy, and it was. * Clean the ICV? Sure, how? Carb cleaner ok? Just spray it in there? * Where can I find diagrams for the fuel lines? Are they even possible to check? It's tight like a... back there. The right side fuel rail was moist after the install, but the engine had been running about 15 mins on the first oil, then about 20 on the second fill, all indoors. We put it down to either condensation, or due to it sitting for a few months and having dried up o-rings and what not. * Ok, I have noticed an increase in fuel consumption, but I put it down to the high idle. The car will try to keep rpms at 1500 so it isn't strange. It also makes the car jump a little when you decrease speed with the car in gear. It will decrease until it hits 1500 and then do a jolt as the car tries to keep the revs. Not good for consumption. I haven't noticed any liquids under the car besides the first few days, and that was glycol (yuck!). It also pulls as strong as ever. All in all, I doubt that I have a fuel leak. * I went over all the smaller vacuum hoses before installing, looked good. Is there an easy way to diagnose? Disconnecting a smaller hose makes no difference on idle, but I haven't tried driving like that. * MY98, so yes, cable-actuated throttle. Checked with an OBDII-reader, the TPS reads 0-100%. It's also properly shut when the gas isn't depressed. No slack, looks good. * Everything looked good on the first startup. Mounts looked good, no excessive movement of the engine, and besides the high idle it's been running fine for the past week and a half. * I've never paid attention to ignition advance, fuel trims or O2 sensor voltages. What readings should I find?

Quotes source post: 9771315

Double-check there is no mechanical reason for the advanced idle. There is a book, 996 Essentials (something like this) that is available that might cover this ICV cleaning operation. I may have cleaned an ICV a long time ago but if so I can't recall and I certainly can't recall how I did it. Diagram for the fuel lines would be in a factory manual. These are not easy or cheap to come by. There used to be one in digital form for downloading at a UK web site, but that site last I read no longer offers the 996 manuals for downloading. A possible source for fuel line diagrams might be say Pelican Parts web site that offers line drawings of the fuel lines as part of its parts listing. You can check fuel pressures and flow which will tell you something about the health of the fuel lines but this kind of work requires someone who is comfortable working with fuel systems. Due to the dangers involved I generally advise non-pro mechanics to farm this work out to a pro. In this area of fuel system checkout/testing/etc. I'm not qualified to advise you beyond this. For vacuum lines, what I used to do was to manipulate vacuum lines with the engine running, and observe if touching/moving any line about affected the engine in any way. But like messing with fuel lines, messing with vacuum lines is dangerous. Get a hand where it shouldn't be and it can get caught in a belt. Even if you are careful, touch something hot, or get a shock from a coil, and you could jerk your hand from the hot or shock source into the serpentine belt. course, long sleeves, watches, rings, any hanging/loose jewelry, etc. belong on the dresser in your bedroom not on your body when working around cars and engines. It has been awhile since I observed ignition advance so I can't say off hand what to look for other than it doesn't want to be too far advanced. IIRC initial ignition advance is around 5 degs. or so. (It can rise to 30 degs. or more advance under some part throttle/moderate to high load conditions but you are not going to create those on your driveway at idle.) Short term fuel trims is tough to describe. They may be positive, maybe close to 10%, upon cold start as the DME is adding extra fuel because the engine is cold and there's more friction, etc. As the engine warms this positive number drops and it can go negative as the DME then as the engine is warmer subtracts fuel. What you mainly are looking at is if one or both sides are acting too erratic or if one side is reading too different from the other side. Oh yeah, for side read bank. O2 sensor voltages at the #1 sensors should vary from under 0.1 volt to around 0.8 volt though often an OBD2 reader can't reads/sample these fast enough so you do not see the peaks and valleys of the readings. The cycle time is maybe close to 2 seconds at (regular idle speed) to less than one second at higher rpms. But this variation in readings comes from the DME adding a bit too much fuel then not adding enough fuel. By this action and the fact the sensors track this the DME knows it has control over fueling. The #2 sensors want to read a fairly steady 0.7 volts or thereabouts. The level varies some as the #1 sensor readings vary but overall the #2 sensors want to show a steady high(er) voltage which is a sign of an absence of oxygen in the exhaust (after the converters) which is a good indication the oxygen is being used in the converters to process exhaust gases. If too much oxygen makes it out of the converters this can be sign they are bad -- they can no longer store oxygen. As good as a place as any: Might as well point out the failure that is causing this advanced idle speed may be a common one and the result is the readings from the various sensors or derived values read the same. A more esoteric explanation of the engine's behavior is there can be at least in some cars an electrical problem in the A/C wiring that erroneously signals the A/C compressor is activated which causes the DME to add more fuel to keep rpms up. If this signal is present and yet the compressor is not activated its parasitic drag is not present and the engine's fast idle can be the result. Absent the proper Porsche diagnostics computer test equipment, one possible check out technique would be to turn on the A/C and see if the engine reacts see if the compressor shows signs it is compressing, then turn off the A/C and observe the engine's reaction. Sincerely, Macster.

I had a similar issue awhile back, turned out the oil fill cap had come loose and was causing a massive vacuum leak.

Quotes source post: 9778759

Had a leaking oil filler tube cap on my Boxster and the idle speed was ok but the CEL came on and from the error codes over time I mistakenly diagnosed the source of the problem as the MAF. Later when the CEL returned a co-worker heard the cap leaking. Replacing the cap made the CEL stay off. But the cap is worth considering in the OP's case. My opinion is that along with the coolant cap the oil filler cap could be considered a wear item and if not replaced as a preventative measure at least monitored for any signs either needs replacement. Unfortunately my experience/knowledge of pre E-gas systems is very limited and I really do not have a good grasp of the ills that can befall engines fitted with pre E-gas throttle/idle control systems. Sincerely, Macster.

Quotes source post: 9779079

Along that same line, it's very easy to crack or dislodge the oil filler tube if you have the engine out of the car and you're swapping out throttle bodies.

Good tips, but I already tried that. I fail to hear any vacuum leaks, and get a huge change in idle if I loosen the oil filler cap. And it's not that hard to remove either so I guess the AOS is still ok

Quotes source post: 9779815

A change in idle is normal. Maybe. I have not had a pre-EGas car to experiment with so I can't say with any confidence what the normal reaction would be/should be. Anyhow, with an EGas car the removal of the oil filler tube cap creates a big intake air leak, unmetered air at that, and the DME can adapt but it takes a few moments. Also, how cold warm or hot the engine is may play a role in the degree the engine reacts. The AOS can have I believe at least two failure modes, or perhaps two stages of its downward spiral to total failure. One is the oil filler tube cap can be unscrewed and removed with the engine running with no real difficulty. This might be in the early stages of an AOS failure. The other is of course just the opposite. In some cases the cap can't be removed at all. This was in the latter stages of an AOS failure, at least that was my impression from when I encountered this oddity. If you are sure the throttle linkage and any electro-mechanical hardware that assists the linkage in controlling the TB and its butterfly valve is ok the dramatic change in idle speed I think is a good clue, but no suspect -- other than those that I have mentioned before -- comes to mind. Let me think some more about this... Have you any means of monitoring intake vacuum/pressure while the engine is running and before, during, and after it exhibits the odd behavior? Sincerely, Macster.

For future reference, the fault was with both o2 sensors. We needed to use a bit of heat and hammerwork to get the exhaust separated and in the process damaged the sensors. Not enough to trigger a CEL though. Not until I did a reset of the ecu and it tried to do its adaption phase. It failed so the car finally threw a CEL and logged error codes for both o2 sensors. Swapped them and all was fine. Thanks for all the help

Wow! Has it only been 6 years? Very impressed you closed this out. Do you still have the car?

Quotes source post: 15599187

haha, no, I sold it may 2013 and got an '02 996 Turbo. In may 2017 I got a 981 Cayman GTS as well, then sold them both in may 2018 and got my current car, a 991.1 Carrera 4 GTS. Here's the 991 GTS and 996 Turbo: